Researchers from the BrainGate consortium have developed a brain-computer interface that converts the brain signals of a paralyzed person into a world record at a rate of 90 characters or 16 words per minute
The BrainGate team revealed that a paralyzed 65-year-old man can successfully use the brain-computer interface and can write 16 words per minute based on his own ideas. This person is the only participant in the study and suffered a spinal cord injury ten years ago that paralyzed his shoulder. The system can successfully translate the signals captured by the electrodes from the human brain, and can write handwritten texts at a speed that breaks the world record with this expertise, twice the speed achieved by previous methods. In addition to the physical components, the interface also uses machine learning algorithms to decode brain signals.
The novelty of this method is that researchers try to capture the mental behavior of calligraphy or handwriting thinking in order to achieve excellent performance in the translation of thought words. Gizmodo. The team placed two 4×4 mm sensors on the outer area of the human brain in the area that controls the movement of the relative parts of the human body, with 100 electrodes, the thickness of which is equivalent to the thickness of hair.
During the experiment, the man tried to move his paralyzed hand to write letters and words, superimposing them on each other and handwriting with a pen. The team used the “greater than” symbol to insert spaces between words. The system can recognize each letter with 95% accuracy. The achieved speed is 16 words per minute, which is equivalent to 75% of the average record level written on mobile phones in the 65-year-old age group (without disabilities).
Jaimie Henderson, head of the Stanford University Neuroprosthesis Translation Laboratory, signed the research and naturalExplains that this novelty will “enable people with speech or movement disabilities to communicate via text, email or other forms of text.” The BrainGate project is led by Frank Willet and supervised by neuroscientists Krishna Shenoy and Henderson. Previous research has made it possible to convert monkey thoughts into words at a rate of 12 words per minute. Subsequent work enabled paralyzed people to “write” at a rate of 40 characters or 8 words per minute.
Although the results of the research are promising, there are still many limitations in the sense that it is an invasive procedure that requires surgery and implants and cannot be applied in a universal manner. The “machine” must learn the nuances of each user, and requires professional technicians to assemble the interface and manage the recognition software.